Channel rate estimation is difficult in an MPEG system with high jitter at the packet input. Due to the high jitter it is not possible to acquire an exact input stream data rate before the data is played out.
Even when the play out rate is accurate there can be a difference in system clock of the stream source (e.g. the upstream video server) and the local system clock (e.g. the EQAM clock) on the order of 30 ppm. As a result, the play out rate can differ slightly from the net input data rate which will eventually lead to de-jitter buffer overflow or buffer underflow.